Task processing method and device based on large model, medium and program product
Through a large-model-based task processing method, integrating the task manager and patent tool library, multi-tool scheduling is realized, solving the problem of inefficiency of traditional patent service models and improving the efficiency and accuracy of patent task processing.
Patent Information
- Application Number
- CN202411131069.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-16
- Publication Date
- 2024-11-22
AI Technical Summary
The traditional patent service model is inefficient and expensive, and cannot provide a one-stop solution. Users need to frequently switch different platforms to process patent tasks, resulting in low efficiency and error-prone.
Using a large model-based task processing method, the big model, task manager and patent tool library are integrated, and through intent identification, goal task determination and tool tag scheduling, corresponding patent service tools are called to realize multi-tool scheduling and task processing.
It improves the efficiency of patent task processing, reduces cross-platform tool switching, reduces manual participation, and realizes the completion of multiple patent service functions in one system, improving processing efficiency and accuracy.
Smart Images

Figure CN119003124A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to a specific field of data processing technology, and in particular to a task processing method, device, medium and program product based on a large model. Background Art
[0002] With the development of science and technology and the enhancement of awareness of intellectual property protection, the application and management of patents have become increasingly important. However, the traditional patent service model is often inefficient and costly, and it is difficult to meet the growing needs of modern enterprises and individuals. Existing patent service tools usually have single functions and cannot provide a one-stop solution. Users need to switch between different platforms and use different patent service tools to process patent tasks, which is time-consuming, error-prone, and inefficient. Therefore, how to improve the processing efficiency of patent task processing has become an urgent problem to be solved.
[0003] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention
[0004] The main purpose of this application is to provide a task processing method, device, medium and program product based on a large model, aiming to solve the technical problem of how to improve the processing efficiency of patent task processing.
[0005] To achieve the above objectives, the present application proposes a task processing method based on a large model, which is applied to a task processing system based on a large model. The task processing system based on a large model deploys a large model, a task manager and a patent tool library, and the patent tool library stores multiple patent service tools. The method includes:
[0006] Obtain the demand information input by the user, perform intent recognition on the demand information according to the pre-trained large model, and obtain the user's intent;
[0007] Determining a target task corresponding to the user intention according to the task manager;
[0008] Determine, according to the task type of the target task and the user intention, a tool label of the patent service tool required to perform the target task;
[0009] The patent service tool corresponding to the tool tag in the patent tool library is called to perform task processing on the target task.
[0010] In one embodiment, the patent service tool includes a patent search tool and a patent interpretation tool, and the tool label includes a patent search label corresponding to the patent search tool and a patent interpretation label corresponding to the patent interpretation tool.
[0011] The step of calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task includes:
[0012] If the target task is a briefing document duplicate checking task, then the patent search tool corresponding to the patent search tag in the patent tool library is called to search the briefing document corresponding to the briefing document duplicate checking task, and obtain and output the first relevant patent document;
[0013] The patent interpretation tool corresponding to the patent interpretation label in the patent tool library is called to perform comparative analysis on the first relevant patent document and the briefing book, and a comparison result of the comparative analysis is output.
[0014] In one embodiment, the patent service tool includes a patent mining tool, a patent search tool and a domain knowledge statistical analysis tool, and the tool label includes a patent mining label corresponding to the patent mining tool, a patent search label corresponding to the patent search tool, and a statistical analysis label corresponding to the domain knowledge statistical analysis tool.
[0015] The step of calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task includes:
[0016] If the target task is a patent creativity mining task, then calling the patent mining tool corresponding to the patent mining tag in the patent tool library to perform patent mining on the user intention, and obtaining and outputting a mining text containing keywords;
[0017] Calling the patent search tool corresponding to the patent search tag in the patent tool library to search for keywords in the mined text to obtain and output a second related patent document;
[0018] The domain knowledge statistical analysis tool corresponding to the statistical analysis tag in the patent tool library is called to extract key information of the user intention, generate a domain knowledge statistical report based on the extracted key information, and output the domain knowledge statistical report.
[0019] In one embodiment, the patent service tool includes a patent drafting tool and a process tracking tool, and the tool tag includes a patent drafting tag corresponding to the patent drafting tool and a process tracking tag corresponding to the process tracking tag.
[0020] The step of calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task includes:
[0021] If the target task is a writing task, the patent writing tool corresponding to the patent writing tag in the patent tool library is called to generate an auxiliary writing copy of the patent briefing based on the user's intention, and output the auxiliary writing copy;
[0022] The process tracking tool corresponding to the process tracking tag in the patent tool library is called to follow up the progress of each patent node corresponding to the writing task, and the progress tracking result is output.
[0023] In one embodiment, after the step of determining the tool label of the patent service tool required to perform the target task, the following steps are included:
[0024] If there are multiple tool tags, determine the execution order of the patent service tools required to perform the target task;
[0025] Determine the scheduling process of the patent service tool corresponding to each tool tag according to the execution order, and execute the step of calling the patent service tool corresponding to the tool tag in the patent tool library according to the scheduling process to perform task processing on the target task.
[0026] In one embodiment, after the step of determining the tool label of the patent service tool required to perform the target task, the method further includes:
[0027] Matching the tool tag with all patent service tools in the patent tool library;
[0028] If there is a matching patent service tool, the matching patent service tool is used as the patent service tool corresponding to the tool tag;
[0029] If there is no matching patent service tool, the data interface of the patent service tool matching the tool tag in the preset external terminal is determined, and the patent service tool matching the tool tag in the external terminal is called according to the data interface to perform task processing on the target task.
[0030] In one embodiment, the method further comprises:
[0031] Acquire a preset training data set, wherein the training data set includes historical patent document data, historical technical document data, and user feedback data;
[0032] Fine-tune the large model according to the historical patent document data and the historical technical document data to obtain a trained large model;
[0033] The patent service tools in the patent tool library are fine-tuned according to the user feedback data to obtain an adjusted patent tool library.
[0034] In addition, to achieve the above purpose, the present application also proposes a task processing device based on a large model, which is arranged in a task processing system based on a large model. The task processing system based on a large model deploys a large model, a task manager and a patent tool library. The patent tool library stores a plurality of patent service tools, including:
[0035] An acquisition module is used to acquire the demand information input by the user, and perform intent recognition on the demand information according to the pre-trained large model to obtain the user's intent;
[0036] A determination module, configured to determine a target task corresponding to the user intention according to the task manager;
[0037] A determination module, used to determine a tool label of a patent service tool required to perform the target task according to the task type of the target task and the user intention;
[0038] The calling module is used to call the patent service tool corresponding to the tool label in the patent tool library to perform task processing on the target task.
[0039] In addition, to achieve the above-mentioned purpose, the present application also proposes an electronic device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the large model-based task processing method as described above.
[0040] In addition, to achieve the above-mentioned purpose, the present application also proposes a medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the large model-based task processing method described above are implemented.
[0041] In addition, to achieve the above-mentioned purpose, the present application also provides a program product, which is a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps of the large model-based task processing method as described above.
[0042] In the embodiment of the present application, a task processing system based on a large model is set up, a large model, a task manager and a patent tool library are deployed in the task processing system based on a large model, and multiple patent service tools are stored in the patent tool library. In addition, multiple patent service tools can be integrated into one system, allowing users to use different patent service tools in one system to achieve different patent service functions, avoiding the need for users to frequently switch between different platforms when they want to achieve different patent service functions at the same time, resulting in low efficiency and prone to errors.
[0043] And it will obtain the demand information input by the user, and perform intent recognition on the demand information according to the pre-trained large model to obtain the user's intention. Then, the model performance of the large model can be used to accurately recognize the demand information input by the user, and obtain the user's intention with relatively high accuracy. According to the task manager, the target task corresponding to the user's intention is determined, and according to the task type and user intention of the target task, the tool label of the patent service tool required to execute the target task is determined, and the patent service tool corresponding to the tool label in the patent tool library is called to perform task processing on the target task. Then, according to the task manager, the target task corresponding to the user's intention is constructed, and then the tool label to be called is determined, and the patent service tool corresponding to the tool label is directly called in the patent tool library, and the task processing of the target task is performed according to the called patent service tool, so that it is possible to achieve the task execution without frequently using different patent service tools across platforms, thereby improving the task processing efficiency, and in the same system, the system itself identifies and calls the task processing tool, and no human participation is required, thereby also improving the processing efficiency of patent task processing. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0045] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0046] Figure 1 A flowchart diagram of the first embodiment of the task processing method based on a large model of the present application;
[0047] Figure 2 A schematic diagram of the system framework of a task processing system based on a large model in the task processing method based on a large model of the present application;
[0048] Figure 3 This is a schematic diagram of the overall process of the task processing method based on a large model in this application;
[0049] Figure 4 This is a schematic diagram of the structure of a task processing device based on a large model according to an embodiment of the present application;
[0050] Figure 5 Schematic diagram of the device structure of the hardware operating environment involved in the large model-based task processing method in the embodiment of the present application.
[0051] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0052] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.
[0053] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.
[0054] The following is a first embodiment of the task processing method based on a large model of the present application. Figure 1 , Figure 1 This is a flow chart of the first embodiment of the task processing method based on a large model of the present application.
[0055] In this embodiment, the task processing method based on the large model is applied to the task processing system based on the large model. Figure 2 ,In the task processing system based on the large model, the task manager and the ,patent tool library are deployed, and multiple patent service tools are stored in the ,patent tool library.
[0056] Optionally, the task processing system based on the large model can be deployed in an electronic device, which can be a computing service device with data processing, network communication and program running functions, such as a server, a tablet computer, a personal computer, a mobile phone, etc. And the electronic device provided with the task processing system based on the large model can be a device deployed in an enterprise.
[0057] Optionally, a large model is also deployed in the task processing system based on the large model. The large model can be a large language model to be trained, that is, a model that can be used to perform natural language processing tasks, such as LLaMa2 13B, LLaMa2 7B, OPT 7B, etc. Optionally, the large model in the task processing system based on the large model needs to be trained in advance to obtain a pre-trained large model that can adapt to various patent service tools in the task processing system based on the large model. During the training process of the large model, the model parameters in the large model can be optimized. Its initial model parameters can be obtained after model pre-training (or the pre-trained model can be used as the model to be trained), or its initial model parameters can also be initialized based on experience or randomly initialized.
[0058] Optionally, the task manager can generate corresponding processing tasks (such as tasks related to patent services) according to the user's intention, so that the task processing system based on the large model can perform the corresponding processing tasks. In addition, the task manager can also perform patent service tool screening operations. Optionally, the task manager can be a management tool in an electronic device, such as a tool with task generation management and patent service tool call, such as a chip, sensor, processor, etc.
[0059] Optionally, the patent tool library can be a database for storing relevant data corresponding to patent service tools. That is, multiple patent service tools are stored in the patent tool library. Optionally, the patent service tool can be a tool for providing different patent services, such as patent search, patent interpretation, patent idea exchange, patent innovation mining, patent knowledge graph generation, domain knowledge statistical analysis and other services, which may include but are not limited to patent search tools, patent interpretation tools, patent writing tools, patent mining tools, patent analysis tools, patent education and consulting tools, patent monitoring and early warning tools, etc. Optionally, patent analysis tools may include knowledge graph generation tools, domain knowledge statistical analysis tools, patent monitoring and early warning tools, etc.
[0060] That is, the task processing system based on the big model in this embodiment integrates the big model, the task manager and the patent tool library containing a variety of patent service tools, such as patent search, patent interpretation, patent idea exchange, patent innovation mining, patent knowledge graph generation, domain knowledge statistical analysis and other services. It can then realize multi-tool scheduling, select and schedule corresponding patent service tools according to user needs, realize intelligent and efficient patent management and services, and effectively improve the efficiency of patent application, management and analysis.
[0061] The task processing method based on a large model includes steps S10 to S40.
[0062] Step S10, obtaining the demand information input by the user, and performing intent recognition on the demand information according to the pre-trained large model to obtain the user's intent;
[0063] In this embodiment, a display device can be set up. The display device can obtain the demand information input by the user, send the demand information to the large model-based task processing system in the electronic device, and display the relevant output data of the large model-based task processing system in the display device so that the user can know the corresponding progress.
[0064] Optionally, the task processing system based on the big model can interact with the user through a display device and provide the user with task progress and results in real time, supporting the collection of user feedback and suggestions. After receiving the user's feedback information, the task processing system based on the big model will analyze the user's feedback information and usage data in real time, and optimize the task processing system based on the big model, such as fine-tuning the big model and fine-tuning each patent service tool.
[0065] Optionally, the display device can set the display page according to the user's usage habits, obtain the user's input demand information in the display page associated with the large model-based task processing system, and send the demand information to the large model-based task processing system (hereinafter referred to as the system). Optionally, when obtaining the user's input demand information, the display device can use different methods such as text input, voice input, and graphical interface input. There is no limitation here. Optionally, the demand information can be text information, voice information and / or image information.
[0066] Optionally, after obtaining the demand information input by the user, the system can identify the intent of the demand information through a pre-trained large model. For example, the user describes the demand through an input box or voice assistant function in the display device, so that the display device can receive the demand information input by the user, such as patent query, patent writing, and patent portfolio analysis. And send the demand information to the system. After receiving the demand information, the system inputs the demand information into the large model to identify the user's intent, so as to understand the user's true intention through the understanding ability of the large model and obtain the user's intention.
[0067] Step S20, determining a target task corresponding to the user's intention according to the task manager;
[0068] Optionally, the large model in the system obtains the user's intention after receiving and identifying the demand information input by the user. The task manager in the system converts the user's intention into a corresponding task definition based on the demand information and the user's intention to obtain the target task.
[0069] For example, if the user intends to search for patents, the task manager will generate a patent search task. If the user intends to write a patent, the task manager will generate a patent writing task.
[0070] Step S30, determining a tool label of a patent service tool required to perform the target task according to the task type of the target task and the user's intention;
[0071] Optionally, a multi-tool scheduling framework is set up in the system so that the scheduling and allocation of patent service tools can be realized through the multi-tool scheduling framework. Optionally, the function of the multi-tool scheduling framework can be realized through a task manager. That is, after generating a target task, the task manager in the system can allocate a suitable patent service tool to the target task, and then call the corresponding patent service tool to execute the target task.
[0072] Optionally, the task type of the target task can be determined, such as whether it is a patent writing task, a patent search task, a patent analysis task, etc. And the tool label of the patent service tool required to perform the target task is determined according to the task type and the user's intention. Optionally, the tool label can include the identification information of the patent service tool, such as the name of the patent service tool, etc., and different identification information can be assigned to different patent service tools, such as 01 for patent search tools, 02 for patent interpretation tools, etc. Different patent service tools correspond to different tool labels, and each patent service tool in the patent tool library has a corresponding tool label.
[0073] Optionally, when determining the tool label according to the task type and user intent, the task type and user intent can be interpreted and identified to determine the patent service tool required to perform the target task and obtain the corresponding tool label. For example, if the task type is a patent writing task, and if the user intent also includes process follow-up, it is determined that the patent service tools required to perform the target task include patent writing tools and process tracking tools. And obtain the tool label corresponding to the patent writing tool, such as the patent writing label. Obtain the tool label corresponding to the process tracking tool, such as the process tracking label.
[0074] In a feasible embodiment, after the step of determining the tool label of the patent service tool required to perform the target task in step S30, steps a10-a20 are included.
[0075] Step a10, if there are multiple tool tags, determine the execution order of the patent service tools required to execute the target task;
[0076] Step a20, determining the scheduling process of the patent service tool corresponding to each tool tag according to the execution order, calling the patent service tool corresponding to the tool tag in the patent tool library according to the scheduling process, and performing task processing on the target task.
[0077] Optionally, if it is determined through the task manager in the system that multiple patent service tools are required to execute the target task and multiple tool tags are obtained, it is necessary to determine the execution order of each patent service tool when the target task is subsequently executed.
[0078] Optionally, after determining the execution order, the scheduling process of the patent service tool corresponding to each tool tag can be determined according to the execution order. For example, if the patent service tool includes a patent search tool, a patent writing tool and a patent mining tool. The execution order is patent mining tool, patent search tool and patent writing tool. Then the scheduling process of the patent service tool corresponding to each tool tag in the patent tool library can be to first schedule the patent mining tool for task processing, then schedule the patent search tool for task processing, and finally schedule the patent writing tool for task processing.
[0079] Optionally, if it is determined through the task manager in the system that only one patent service tool is needed to execute the target task, the patent service tool can be directly called in the patent tool library to process the target task.
[0080] In this embodiment, when it is determined that there are multiple tool tags, the execution order corresponding to the patent service tools required to execute the target task is determined, and then the scheduling process is determined according to the execution order. According to the scheduling process, the patent service tools corresponding to the tool tags in the patent tool library are called in sequence to perform task processing, thereby ensuring the effective implementation of task processing and improving task processing efficiency.
[0081] In a feasible embodiment, after the step S30 of determining the tool label of the patent service tool required to perform the target task, steps b10-b30 are also included.
[0082] Step b10, matching the tool tag with all patent service tools in the patent tool library;
[0083] Step b20: if there is a matching patent service tool, the matching patent service tool is used as the patent service tool corresponding to the tool tag;
[0084] Step b30, if there is no matching patent service tool, determine the data interface of the patent service tool that matches the tool tag in the preset external terminal, and call the patent service tool that matches the tool tag in the external terminal according to the data interface to perform task processing on the target task.
[0085] Optionally, if the task manager in the system determines that multiple patent service tools are required to execute the target task, and multiple tool tags are obtained, each tool tag can be matched with all patent service tools in the patent tool library to determine the patent service tool corresponding to each tool tag. This facilitates the subsequent calling of the patent service tool corresponding to the tool tag in the patent tool library to process the target task.
[0086] Optionally, for each tool tag, the tool tag data of each patent service tool stored in the patent tool library can be determined, and the tool tag is matched with each stored tool tag data. If there is matching tool tag data, the patent service tool corresponding to the matching tool tag data can be used as the patent service tool corresponding to the tool tag. Optionally, after determining the matching tool tag data, it can be determined whether the status of the patent service tool corresponding to the matching tool tag data is abnormal. If there is no abnormality, the patent service tool corresponding to the matching tool tag data is used as the patent service tool corresponding to the tool tag. If there is an abnormality, the tool tag data is skipped, and a prompt message indicating the abnormality of the patent service tool can be output, and the tool tag continues to be matched with the remaining tool tag data.
[0087] Optionally, if there is no tool tag data matching the tool tag, it is determined that an external patent service tool needs to be called so as to perform the target task according to the called external patent service tool. Optionally, the tool parameter information of the patent service tool corresponding to the tool tag can be determined first. And according to the tool parameter information, a query can be made online to determine the external terminal (such as a server of another company) containing the patent service tool matching the tool tag, and an access request can be sent to the external terminal, and after receiving the authorization license fed back by the external terminal, the data interface of the patent service tool can be obtained. So as to call the patent service tool matching the tool tag in the external terminal through the data interface, perform task processing on the target task, obtain the task processing result, and output the task processing result.
[0088] Optionally, if it is determined through the task manager in the system that a patent service tool is required to execute the target task, that is, only a tool tag is obtained, this tool tag can be directly matched with each patent service tool in the patent service tool library to obtain the corresponding patent service tool.
[0089] In this embodiment, the tool label is matched with all patent service tools in the patent tool library, and when a match is determined, the matching patent service tool in the patent tool library is directly determined to be the patent service tool corresponding to the tool label. When a mismatch is determined, the patent service tool matching the tool label in the external terminal is called according to the data interface of the external terminal to perform task processing on the target task, thereby ensuring the effective implementation of task processing and improving the efficiency of task processing.
[0090] Step S40, calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task.
[0091] Optionally, after determining the patent service tool corresponding to the tool tag in the patent tool library, the task manager in the system can directly call the patent service tool to execute the target task. Optionally, when there are multiple patent service tools that need to be called, the patent service tools in the patent tool library can be called in sequence according to the scheduling process, and the target task can be processed in sequence according to each called patent service tool.
[0092] In addition, to assist in understanding the overall processing flow of the task processing system based on the large model in this embodiment, an example is given below.
[0093] For example, Figure 3 As shown, the task processing system based on the large model obtains user input information, and identifies the user's intention through the large model, and then determines the target task corresponding to the identified user's intention through the task manager, and determines the tool label of the patent service tool required to execute the target task according to the task type and user's intention of the target task, and then calls the patent service tool corresponding to the tool label in the AI Agent (artificial intelligence) tool (such as the patent tool library) through the tool scheduling center to perform task processing of the target task, such as selecting or combining tools to process tasks, and outputs the processing results of the task processing to the user. Among them, a variety of patent service tools can be set in the patent tool library, such as patent search, patent interpretation, patent writing, patent mining, statistical analysis, patent education and consultation, patent monitoring and early warning and other patent service tools.
[0094] In this embodiment, a task processing system based on a large model is set up, a large model, a task manager and a patent tool library are deployed in the task processing system based on a large model, and multiple patent service tools are stored in the patent tool library. It is then possible to integrate multiple patent service tools into one system, allowing users to use different patent service tools in one system to achieve different patent service functions, avoiding the need for users to frequently switch between different platforms when they want to achieve different patent service functions at the same time, resulting in low efficiency and prone to errors.
[0095] And it will obtain the demand information input by the user, and perform intent recognition on the demand information according to the pre-trained large model to obtain the user's intention. Then, the model performance of the large model can be used to accurately recognize the demand information input by the user, and obtain the user's intention with relatively high accuracy. According to the task manager, the target task corresponding to the user's intention is determined, and according to the task type and user intention of the target task, the tool label of the patent service tool required to execute the target task is determined, and the patent service tool corresponding to the tool label in the patent tool library is called to perform task processing on the target task. Then, according to the task manager, the target task corresponding to the user's intention is constructed, and then the tool label to be called is determined, and the patent service tool corresponding to the tool label is directly called in the patent tool library, and the task processing of the target task is performed according to the called patent service tool, so that it is possible to achieve the task execution without frequently using different patent service tools across platforms, thereby improving the task processing efficiency, and in the same system, the system itself identifies and calls the task processing tool, and no human participation is required, thereby also improving the processing efficiency of patent task processing.
[0096] Based on the first embodiment of the present application, a second embodiment of the present application is proposed. In the second embodiment of the present application, the same or similar contents as those of the above-mentioned first embodiment can be referred to the above introduction and will not be repeated later.
[0097] In a feasible embodiment, the patent service tool includes a patent search tool and a patent interpretation tool, and the tool label includes a patent search label corresponding to the patent search tool and a patent interpretation label corresponding to the patent interpretation tool. Optionally, the patent search label may include identification information of the patent search tool, such as the name of the patent search tool, etc. The patent interpretation label may include identification information of the patent interpretation tool, such as the name of the patent interpretation tool, etc.
[0098] Step S40, calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task, includes steps c10-c20.
[0099] Step c10: if the target task is a briefing document duplicate checking task, then call the patent search tool corresponding to the patent search tag in the patent tool library to search the briefing document corresponding to the briefing document duplicate checking task, and obtain and output the first relevant patent document;
[0100] Optionally, if the user intention identified by the large model is that the user needs to check the duplication of his own briefing document (i.e., perform a novelty search) and compare the similarities and differences with the queried patent documents, then the target task generated by the task manager can be determined as the briefing document duplication check task.
[0101] Optionally, in the patent tool library, the patent search tool corresponding to the patent search tag can be called, and the patent database can be searched through the patent search tool to return a list of patents matching the disclosure document. The patent list will display the highly relevant patents found and their similarities. This patent list and its corresponding patents can be used as the first relevant patent document.
[0102] Optionally, the patent search tool may include a patent database and a query engine. The patent database may store patent data from different patent offices, such as patent title, abstract, full text, classification number, applicant, inventor, legal status, etc., as well as abstract, claims, specification, drawings, etc.
[0103] Optionally, the query engine may include two parts, namely indexing and retrieval, and natural language processing. For the indexing and retrieval part, search engine technologies such as Elasticsearch (search server) or Solr (enterprise-level search application server) may be used to index the patent database, supporting efficient keyword retrieval, advanced query, semantic query and other functions. Optionally, for the natural language processing part, NLP technology may be combined to support users to query in natural language and convert the query into a structured search request.
[0104] Optionally, the input of the patent search tool can be the user's problem description (that is, it can be determined based on the user's intention identified by the large model) or the keywords and text description used by the user for query (such as the keywords and text description information contained in the briefing document uploaded by the user). The output of the patent search tool can be the patents retrieved by the patent search tool and the similarity ranking, and can be output in the form of a patent list.
[0105] Step c20, calling the patent interpretation tool corresponding to the patent interpretation label in the patent tool library, performing comparative analysis on the first relevant patent document and the briefing document, and outputting the comparative analysis result.
[0106] Optionally, in the patent tool library, the patent interpretation tool corresponding to the patent interpretation label can be called to compare and analyze the first relevant patent document retrieved by the patent search tool with the disclosure document. That is, the similarity and difference between each patent in the patent list and the disclosure document are analyzed, and the comparative analysis results are generated and output, and then the user is assisted in judging whether the creativity of the disclosure document is sufficiently innovative and whether the technical solution of the disclosure document is covered by existing patents based on the comparative analysis results.
[0107] Optionally, the patent interpretation tool may include a natural language processing model and a professional terminology explanation. Optionally, the natural language processing model may perform functions such as text summary generation, key point extraction, and content comparison. For example, through the text summary generation model, key content is extracted from the full text of the patent to generate a concise and easy-to-understand patent summary. For example, using information extraction technology, key technical points, innovations, and technical effects in the patent are identified and extracted. For example, combined with the natural language processing large model technology, the full text of the patent is understood, and the models are compared to compare the semantic similarities and differences.
[0108] Optionally, for the professional terminology interpretation part, a terminology library and a terminology parsing module are set in the patent interpretation tool to realize the professional terminology interpretation. The terminology library includes professional terms and their interpretations in patents and different technical fields. The terminology parsing module can support the interpretation of professional terms in patent texts to help users understand the patent content.
[0109] Optionally, the input part of the patent interpretation tool can be the user's problem description (such as the user's intention identified by the big model), the full text of the patent (such as the full text of each patent in the patent list retrieved by the patent search tool), the full text of the briefing document (such as the briefing document uploaded by the user), etc. The output part of the patent interpretation tool can be the text content of the answer, the result of the patent interpretation, such as the extraction and overview of key technical points, the overview of technical effects and advantages, etc. It can also be the comparison results between patents, the comparison results between patents and briefing documents, and the comparison results between patents and creativity. The comparison results can include similarities and differences in technical innovation points, similarities and differences in technical effects and advantages.
[0110] In this embodiment, when it is determined that the target task is the briefing document duplication check task, the patent search tool in the patent tool library is called to search the briefing document, and the first relevant patent document is obtained and output, and then the patent interpretation tool in the patent tool library is called to compare and analyze the first relevant patent document and the briefing document, and the comparison result of the comparison analysis is output, and then the briefing document duplication check task is completed, so that two tools are called in one system without the user having to switch between different platforms, thereby improving the processing efficiency of the target task.
[0111] In a feasible embodiment, the patent service tool includes a patent mining tool, a patent search tool and a domain knowledge statistical analysis tool, and the tool label includes a patent mining label corresponding to the patent mining tool, a patent search label corresponding to the patent search tool, and a statistical analysis label corresponding to the domain knowledge statistical analysis tool. Optionally, the patent mining label may include identification information of the patent mining tool, such as the name of the patent mining tool, etc. The statistical analysis label may include identification information of the domain knowledge statistical analysis tool, such as the name of the domain knowledge statistical analysis tool, etc.
[0112] Step S40, calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task, includes steps d10-d30.
[0113] Step d10: If the target task is a patent creativity mining task, then call the patent mining tool corresponding to the patent mining tag in the patent tool library to perform patent mining on the user's intention, and obtain and output the mining text containing the keyword;
[0114] Optionally, if the user intention identified by the large model is that the user needs to generate creative ideas with the help of the system, then the target task generated by the task manager can be determined to be a patent creativity mining task.
[0115] Optionally, the patent mining tool corresponding to the patent mining tag can be called in the patent tool library, and the patent mining tool can be used to communicate and generate answers. For example, the user intention identified by the large model can be mined for keywords to generate mining text containing keywords (such as technical keywords).
[0116] Optionally, the patent mining tool may include data mining and machine learning, as well as innovation assessment. Optionally, for the data mining and machine learning part, innovation point identification can be performed, such as using machine learning and data mining technology to automatically identify potential innovation points and technology trends from a large amount of patent data. It is also possible to generate a technology roadmap, which shows the course of technology development and possible future innovation directions. For the innovation assessment part, the evaluation model can be used to analyze the technical novelty, creativity and market potential of the patent. In addition, the market data and technology development trends can be combined to conduct risk analysis on the innovation points and provide decision support.
[0117] Optionally, the input of the patent mining tool can be the user's question description, and the output can be the text content of the answer and the drawn technology roadmap. The output text content of the answer and the technology roadmap can be used as mining text containing keywords.
[0118] Step d20, calling the patent search tool corresponding to the patent search tag in the patent tool library, searching for the keywords in the mined text, and obtaining and outputting the second relevant patent document;
[0119] Optionally, the patent search tool corresponding to the patent search tag can be called in the patent tool library, and the patent database can be searched through the patent search tool to search for keywords in the mined text to obtain a list of patents matching the keywords. The highly relevant patents found and their similarities will be displayed in the list of patents matching the keywords. The patent list matching the keywords and the corresponding patents can be used as the second relevant patent document.
[0120] Step d30, calling the domain knowledge statistical analysis tool corresponding to the statistical analysis tag in the patent tool library, extracting key information of the user's intention, generating a domain knowledge statistical report based on the extracted key information, and outputting the domain knowledge statistical report.
[0121] Optionally, in the patent tool library, the domain knowledge statistical analysis tool corresponding to the statistical analysis tag can be called to extract key information based on the question or description entered by the user, and generate a domain knowledge statistical report related to the user input content. For example, key information is extracted based on the user intent identified by the large model. And according to the user's customized requirements, a statistical analysis report related to the extracted key information can be output, and the technical trend analysis and patent coverage of the technical direction can be analyzed from the report. In addition, data visualization result charts can be output for users.
[0122] Optionally, the domain knowledge statistical analysis tool may include two parts: data analysis and report generation. For the data analysis part, data statistics may be performed, such as using statistical analysis tools to perform statistical analysis on patent data (such as the second relevant patent document) to extract technical hotspots, innovation trends and patent layouts in the field. A data visualization module may also be set up to generate charts (such as bar charts, pie charts and heat maps, etc.) for display. Optionally, for the report generation part, automatic report generation may be performed, that is, a domain knowledge statistical report may be generated based on the statistical analysis results. The domain knowledge statistical report may include technical trend analysis, changes in the number of patent applications, and analysis of major competitors, etc.
[0123] Optionally, the input of the domain knowledge statistical analysis tool can be user intent (such as user problem description and customized report content), and the output can be text content of the answer and data visualization result charts, etc.
[0124] In this embodiment, when it is determined that the target task is a creative mining task, the patent mining tool in the patent tool library is called to perform patent mining on the user's intention, and a mining text containing keywords is obtained and output, and the patent search tool in the patent tool library is called to perform keyword search to obtain and output the second relevant patent document, and then the domain knowledge statistical analysis tool in the patent tool library is called to generate a domain knowledge statistical report and output it, thereby realizing the processing of the two branches of technical documents and patent documents, so that the completion effect of the patent creative mining task is more accurate.
[0125] In a feasible embodiment, the patent service tool includes a patent drafting tool and a process tracking tool, and the tool tag includes a patent drafting tag corresponding to the patent drafting tool and a process tracking tag corresponding to the process tracking tag. Optionally, the patent drafting tag may include identification information of the patent drafting tool, such as the name of the patent drafting tool, etc. The process tracking tag may include identification information of the process tracking tool, such as the name of the process tracking tool, etc.
[0126] Step S40, calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task, includes steps e10-e20.
[0127] Step e10: if the target task is a writing task, then call the patent writing tool corresponding to the patent writing tag in the patent tool library, generate an auxiliary writing copy of the patent briefing based on the user's intention, and output the auxiliary writing copy;
[0128] Optionally, if the user intention identified by the large model is that the user needs to use the system to write a patent briefing document, then the target task generated by the task manager can be determined as a writing task.
[0129] Optionally, the patent writing tool corresponding to the patent writing tag can be called in the patent tool library, and the called patent writing tool can interact with the user. That is, the user inputs a question, which is recognized by the large model, and then the corresponding answer is generated by the patent writing tool, and the answer is used as an auxiliary writing copy and output. Optionally, the patent writing tool can also generate a chain of thoughts corresponding to the answer, and output the chain of thoughts and the answer as an auxiliary writing copy to assist the user in writing the patent briefing document.
[0130] Optionally, in one scenario, the patent writing tool can be called to communicate based on the questions or descriptions input by the user, and the corresponding answers (such as the generated auxiliary writing copy) can be generated to assist the user in writing the patent briefing. Optionally, the user inputs the question into the system, the large model in the system recognizes it, and calls the patent writing tool to answer the recognized question, generate the corresponding reply text, and output the reply text to assist the user in writing the patent briefing.
[0131] Step e20, calling the process tracking tool corresponding to the process tracking tag in the patent tool library, tracking the progress of each patent node corresponding to the writing task, and outputting the progress tracking result.
[0132] Optionally, the process tracking tool can track the progress of patent writing and patent application, and can provide prompts when the process enters key time nodes.
[0133] Optionally, the process tracking tool corresponding to the process tracking tag can be called in the patent tool library, and the progress of different patent nodes can be followed up through the called process tracking tool. Optionally, the patent node can include a process node, a patent document writing node, an inventor node, a handler node, a leadership node, etc. After the inventor completes the patent briefing document with the assistance of the patent writing tool, he can enter the patent document writing node from the inventor node. At this time, the process tracking tool can be used to prompt the user who is in the patent document writing node.
[0134] In this embodiment, when the target task is a writing task, the patent writing tool in the patent tool library is called to generate and output auxiliary writing documents to assist users in writing patent briefing documents, and the process tracking tool in the patent tool library is called to follow up on the progress of each patent node corresponding to the writing task and output the corresponding results. This allows users to use patent writing tools and process tracking tools without switching between different platforms, thereby improving the processing efficiency of writing tasks.
[0135] Furthermore, in a feasible embodiment, the method for task processing based on a large model also includes steps x10-x30.
[0136] Step x10, obtaining a preset training data set, wherein the training data set includes historical patent document data, historical technical document data and user feedback data;
[0137] Step x20, fine-tuning the preset large model according to the historical patent document data and the historical technical document data to obtain a trained large model;
[0138] Step x30, fine-tune the patent service tools in the patent tool library based on user feedback data to obtain an adjusted patent tool library.
[0139] Optionally, in this embodiment, it is necessary to adjust the big model in the system and each patent service tool in the patent tool library to make the performance of the big model and each patent service tool in the system better.
[0140] Optionally, data collection may be performed in advance to construct a training data set based on the collected data. Optionally, the collected data may include historical patent document data, historical technical document data, and user feedback data.
[0141] Optionally, the historical patent document data may be the patent document data that has been disclosed before the current moment, including patent applications, examination opinions, authorized patents and related legal documents. The historical technical document data may be the technical document data that has been disclosed before the current moment, including technical documents, research papers and industry reports in different fields. The user feedback data may be the input data and feedback information of the user when using various patent service tools in the task processing system based on the large model.
[0142] Optionally, after obtaining the training data set, the historical patent document data, historical technical document data and user feedback data contained in the training data set can be cleaned and annotated. For example, the training data set can be cleaned to remove irrelevant content, duplicate content and noise data to ensure the data quality in the training data set. The data in the training data set can be annotated, including annotating the structure of the patent document (such as background technology, invention content, claims, etc.), technical fields, key technical points and innovation points, etc.
[0143] Optionally, the historical patent document data and the historical technical document data can be input into the big model to fine-tune the model parameters in the preset big model to obtain a trained big model. Optionally, the preset big model can be a big model that has been trained in advance.
[0144] Optionally, user feedback data can also be determined, and the patent service tools in the patent tool library can be fine-tuned according to the user feedback data to obtain adjusted patent service tools. At this time, the patent tool library is an adjusted patent tool library. For example, if the user feedback data includes that the performance of the patent search tool is low, the patent search tool in the patent tool library can be fine-tuned according to the user feedback data, so that the performance of the patent search tool is improved. Optionally, supervised learning technology can be used to fine-tune the data annotated in the training data set, so that the large model can learn different capabilities and functions according to the task requirements of different patent service tools.
[0145] In this embodiment, by obtaining a preset training data set, the big model is fine-tuned according to the historical patent document data and historical technical document data in the training data set to obtain a trained big model, and the patent service tool is fine-tuned according to the user feedback data to obtain an adjusted patent tool library, and then an adjusted task processing system based on the big model is obtained, thereby improving the system performance of the task processing system based on the big model.
[0146] In addition, to achieve the above purpose, refer to Figure 4The present application also proposes a task processing device based on a large model, setting a task processing system based on a large model, deploying a large model, a task manager and a patent tool library in the task processing system based on a large model, and storing a plurality of patent service tools in the patent tool library, including:
[0147] The acquisition module A10 is used to obtain the demand information input by the user, perform intent recognition on the demand information according to the pre-trained large model, and obtain the user's intent;
[0148] A determination module A20, used to determine a target task corresponding to the user's intention according to the task manager;
[0149] A determination module A20, for determining a tool label of a patent service tool required to perform the target task according to the task type of the target task and the user's intention;
[0150] The calling module A30 is used to call the patent service tool corresponding to the tool label in the patent tool library to perform task processing on the target task.
[0151] The task processing device based on a large model provided by the present application adopts the task processing method based on a large model in the above-mentioned embodiment, which can improve the processing efficiency of the patent task processing. Compared with the prior art, the beneficial effects of the task processing device based on a large model provided by the present application are the same as the beneficial effects of the task processing method based on a large model provided by the above-mentioned embodiment, and the other technical features in the task processing device based on a large model are the same as the features disclosed in the above-mentioned embodiment method, which will not be repeated here.
[0152] The present application provides a task processing device based on a large model, and the task processing device based on a large model includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the task processing method based on the large model in the above-mentioned embodiment one.
[0153] Reference below Figure 5, which shows a schematic diagram of the structure of a large-model-based task processing device suitable for implementing the embodiment of the present application. The large-model-based task processing device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.
[0154] like Figure 5 As shown, the task processing device based on the large model may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for device operation are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the large model-based task processing device to communicate with other devices wirelessly or by wire to exchange data. Although the diagram shows a large model-based task processing device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.
[0155] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.
[0156] The task processing device based on a large model provided by the present application adopts the task processing method based on a large model in the above-mentioned embodiment, which can improve the processing efficiency of the patent task processing. Compared with the prior art, the beneficial effects of the task processing device based on a large model provided by the present application are the same as the beneficial effects of the task processing method based on a large model provided by the above-mentioned embodiment, and the other technical features in the task processing device based on a large model are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.
[0157] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.
[0158] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
[0159] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, wherein the computer-readable program instructions are used to execute the large model-based task processing method in the above-mentioned embodiment.
[0160] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.
[0161] The computer-readable storage medium may be included in the task processing device based on the large model; or may exist independently without being assembled into the task processing device based on the large model.
[0162] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by a task processing device based on a large model, the task processing device based on the large model can execute the step flow of the task processing method based on the large model.
[0163] Computer program code for performing the operations of the present application may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0164] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0165] The modules involved in the embodiments of the present application may be implemented by software or hardware, wherein the name of the module does not limit the unit itself in some cases.
[0166] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned large-model-based task processing method, and can improve the processing efficiency of patent task processing. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the large-model-based task processing method provided by the above-mentioned embodiment, and will not be repeated here.
[0167] The present application also provides a computer program product, including a computer program, which implements the steps of the large model-based task processing method as described above when the computer program is executed by a processor.
[0168] The computer program product provided by this application can improve the processing efficiency of patent task processing. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the task processing method based on the large model provided by the above embodiment, which will not be repeated here.
[0169] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.
Claims
1. A task processing method based on a large model, characterized in that: Applied to a task processing system based on a large model, wherein the task processing system based on the large model deploys a large model, a task manager and a patent tool library, wherein the patent tool library stores multiple patent service tools, the method includes: Obtain the demand information input by the user, perform intent recognition on the demand information based on the pre-trained large model, and obtain the user's intention; determining a target task corresponding to the user intention according to the task manager; Determining, based on the task type of the target task and the user's intention, a tool tag of a patent service tool required to perform the target task; The patent service tool corresponding to the tool tag in the patent tool library is called to perform task processing on the target task.
2. The method according to claim 1, wherein The patent service tool includes a patent search tool and a patent interpretation tool, and the tool label includes a patent search label corresponding to the patent search tool and a patent interpretation label corresponding to the patent interpretation tool. The step of calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task includes: If the target task is a briefing document duplicate checking task, then calling the patent search tool corresponding to the patent search tag in the patent tool library to search the briefing document corresponding to the briefing document duplicate checking task, and obtaining and outputting the first relevant patent document; The patent interpretation tool corresponding to the patent interpretation label in the patent tool library is called to perform a comparative analysis on the first relevant patent document and the briefing document, and a comparison result of the comparative analysis is output.
3. The method according to claim 1, wherein The patent service tool includes a patent mining tool, a patent search tool and a domain knowledge statistical analysis tool. The tool tags include a patent mining tag corresponding to the patent mining tool, a patent search tag corresponding to the patent search tool, and a statistical analysis tag corresponding to the domain knowledge statistical analysis tool. The step of calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task includes: If the target task is a patent creativity mining task, then calling the patent mining tool corresponding to the patent mining tag in the patent tool library to perform patent mining on the user intention, and obtaining and outputting a mining text containing keywords; Calling a patent search tool corresponding to the patent search tag in the patent tool library to search for keywords in the mined text to obtain and output a second relevant patent document; The domain knowledge statistical analysis tool corresponding to the statistical analysis tag in the patent tool library is called to extract key information of the user intention, generate a domain knowledge statistical report based on the extracted key information, and output the domain knowledge statistical report.
4. The method according to claim 1, wherein The patent service tool includes a patent drafting tool and a process tracking tool, and the tool tag includes a patent drafting tag corresponding to the patent drafting tool and a process tracking tag corresponding to the process tracking tag. The step of calling the patent service tool corresponding to the tool tag in the patent tool library to perform task processing on the target task includes: If the target task is a writing task, the patent writing tool corresponding to the patent writing tag in the patent tool library is called to generate an auxiliary writing copy of the patent briefing document based on the user intention, and output the auxiliary writing copy; The process tracking tool corresponding to the process tracking tag in the patent tool library is called to follow up the progress of each patent node corresponding to the writing task, and output the progress tracking result.
5. The method according to claim 1, wherein After the step of determining the tool label of the patent service tool required to perform the target task, the method includes: If there are multiple tool tags, determine the execution order of the patent service tools required to perform the target task; Determine the scheduling process of the patent service tool corresponding to each tool tag according to the execution order, and execute the step of calling the patent service tool corresponding to the tool tag in the patent tool library according to the scheduling process to perform task processing on the target task.
6. The method according to claim 1, wherein After the step of determining the tool label of the patent service tool required to perform the target task, the method further includes: Matching the tool tag with all patent service tools in the patent tool library; If there is a matching patent service tool, the matching patent service tool is used as the patent service tool corresponding to the tool tag; If there is no matching patent service tool, the data interface of the patent service tool that matches the tool tag in the preset external terminal is determined, and the patent service tool that matches the tool tag in the external terminal is called according to the data interface to perform task processing on the target task.
7. The method according to any one of claims 1 to 6, wherein: The method further comprises: Acquiring a preset training data set, wherein the training data set includes historical patent document data, historical technical document data, and user feedback data; Fine-tuning the large model based on the historical patent document data and the historical technical document data to obtain a trained large model; The patent service tools in the patent tool library are fine-tuned according to the user feedback data to obtain an adjusted patent tool library.
8. An electronic device, characterized in that: The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the large model-based task processing method according to any one of claims 1 to 7.
9. A medium, characterized in that The medium is a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the large model-based task processing method according to any one of claims 1 to 7 are implemented.
10. A program product, characterized in that The program product is a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the large model-based task processing method according to any one of claims 1 to 7 are implemented.
Citation Information
Cited By
Intellectual property software development sharing management method and system
CN120632835A
Data processing method and device based on intention, equipment, medium and product
CN120725182A
OA reply method based on AI agent of intellectual property application process management system
CN122489726A
